College - Author 1

College of Science and Mathematics

Department - Author 1

Physics Department

Degree Name - Author 1

BS in Physics



Primary Advisor

Hilary Jacks, College of Science and Mathematics, Physics Department


A system has the ability to store memory if one is able to write, retrieve, and erase information from it. Some systems are capable of storing multiple transient memories; with no noise in these systems, at long time, the memories degrade until one or two remain. The addition of noise to these systems can extend the retention of multiple memories, in some cases indefinitely [2, 3]. While this behavior was first observed in simulations of charge density waves, it has since appeared in other systems [1].

Past research has shown that sheared non-Brownian liquid suspensions of particles exhibit similar behavior with regards to multiple memory storage with noise [2, 3, 5]. In these systems, a memory of the driving shear that a system has been trained with can be written. Training consists of many cycles of driving at a given strain amplitude γ. After training, there is a sharp increase in irreversibility observed for shearing with a strain amplitude of γ ′ > γ [2]. As a result, we can say that the system has formed a memory at γ.

In order to write multiple memories in a system, multiple amplitudes are used for the training in some predetermined pattern along with noise. Suppose we intended to train a system with two memories, γ1 and γ2 where γ1 > γ2. In order to train the system to have memories at both amplitudes we may repeat a pattern such as γ1, γ2, γ2, γ2, γ2. The system is sheared to each amplitude successively, with noise applied between each shear, and then the pattern repeats. After sufficient repetitions of this pattern, multiple memories are detectable and persist at long time [3, 4].

While so far we’ve focused on sheared systems, past research by Keim, Paulsen, and Nagel indicates that multiple transient memory behavior is not sensitive to the deformation geometry of the system; for example, a system of swelling particles exhibits similar behavior [3]. This is the type of system used in this study. In this case, training the system is not done by shearing the suspension of particles to a given amplitude or pattern of amplitudes. Instead, the system is trained by swelling particles to a specified size or repeated pattern of sizes.

In these swelling systems, we investigate what system parameters influence memory capacity. Memory capacity can be defined as the total number of memories that can be written, stored, and read within one system at the same time. Many parameters about a given system, e.g. the amount of noise, the interaction strength, and the number of particles, may affect the memory capacity.